United Kingdom
I design and build AI products for expert users. My work focuses on tools that help professionals make sense of complex systems, from lawyers reviewing contracts and scientists exploring biomedical evidence to engineers developing AI models. Across legal tech, biotech, and engineering, I've designed and shipped products that translate advanced AI capabilities into structured workflows and practical applications. My experience includes human-in-the-loop agentic systems, knowledge graph interfaces, data dashboards, ML pipeline tools, and deep learning model training platforms. I specialise in product design, UX engineering, design thinking, and AI systems thinking. I'm particularly interested in designing AI tools that experts can trust, inspect, control, and seamlessly integrate into their existing workflows.
• Design and build legal AI products, Skills, plugins, and bespoke applications used by lawyers across a global law firm. • Lead UX design and front-end development from early concept through production delivery. • Design agentic and LLM-powered workflows for legal use cases, including contract analysis, structured review, and human-in-the-loop evaluation. • Translate advanced AI capabilities into trustworthy, scalable legal products with clear user control and transparency. • Rapidly prototype solutions using AI-assisted design and development workflows. • Partner with lawyers, engineers, researchers, and business stakeholders to align product strategy with real legal workflows and commercial needs.
• Designed and built applications for developing, training, and analysing deep learning models in MATLAB. • Led the design of Deep Network Designer, enabling engineers to visually construct, inspect, and edit neural network architectures. • Designed structured app workflows for time-series modelling, including model training, evaluation, compression, and iterative refinement. • Led workshops that translated complex machine learning concepts into intuitive, interactive tools for engineers and technical users. • Defined reusable design system components for deep learning workflows. • Led a small research and design team to explore future product directions, contributing to the product roadmap and delivery of next-generation toolbox features. • Collaborated closely with engineers and researchers to align the product experience with technical requirements, model behaviour, and user needs.
• Designed and built AI-powered tools for biomedical research, helping scientists discover and evaluate drug–disease relationships. • Created interfaces for exploring knowledge graphs, large-scale biological datasets, and model-derived evidence. • Designed “Evidence Chains” to make AI-derived relationships explainable and actionable for scientific users. • Helped reduce research and curation workflows from days to hours by improving how evidence was surfaced, structured, and reviewed. • Partnered with pharmacologists, bioinformaticians, machine learning teams, and product leaders to align tools with real scientific workflows. • Conducted user research, usability testing, and iterative prototyping to validate product direction and improve adoption.
• Designed products for commercial insurance underwriting across API and GUI-based workflows. • Translated risk models and structured data into clear, usable tools for underwriters and technical users. • Designed low-code interfaces that enabled underwriters to assess, configure, and integrate risk data into existing workflows. • Designed and shipped an API documentation platform to help developers understand, consume, and integrate risk data APIs. • Conducted user research and rapid prototyping to validate early-stage product direction and refine market fit.
• Designed and built data exploration, ML pipeline, and visualisation interfaces for AI products, including Kedro and Vizro, which are now open source and used by consultants and data scientists across client engagements. • Led the design of QB1, an AI toolkit that helped standardise machine learning workflows across McKinsey QuantumBlack AI projects. • Worked embedded with data scientists and engineers to translate complex technical processes into structured, usable interfaces, closing the gap between model behaviour and user comprehension. • Established reusable design patterns and component logic across AI tools, improving consistency and reducing ramp-up time on new client projects.